You buy a server to provide a service, thus we measure our success on that server being up. Our target is for your servers to have a combined uptime exceeding 99.9% every month (higher if the server is designed for it). Where a server isn't designed to support that level of uptime we'll provide specific, actionable advice on what could be done.

As for user requests they want their issues resolved quickly, thus we track the time from opening to resolution. Our target is to resolve all issues within one business day, unless parts need to be ordered, you have internal approval processes that must be done first or they request otherwise.

.

Now isn't that a whole lot better?